Paralleled EZ81 - wiring?

[G6Tanuki]

A friend whom is 'into' vintage instrument-amps is building one that will use a pair of 6L6; he wants to use valves as the rectifiers too, given that at full-whang the amplifier will be pulling around 220mA I have offered him a known-good pair of Philips 'Miniwatt' EZ81 which should handle the demand without difficulty.

Question is, should we do the traditional thing and strap each valve as a single half-wave rectifier [anodes commoned through a pair of low-value - 150 Ohm? resistors to prevent one half hogging the current?]

or

I have an alternative suggestion - which saves two resistors [at the expense of the remaining two resistors needing to be double the dissipation].

Wire the resistors in series with the cathodes, and then cross-connect the anodes so there is one anode in each valve connected to the ends of the HT transformer.

Either way, if you do a Thevenin-esque 'route map' of the current-flow, each valve's anode always has a resistor in series with it to limit the current it passes and balance the current against its other conducting anode in the other valve.

Whatever option we go for, we will use rather well-rated balancing resistors - probably the big aluminium-clad bolt-to-a-heatsink wirewounds; I've seen what happens when a limiting resistor in series with one half of a rectifier decides to go open-circuit but the band plays on....

[Leon Crampin]

Given that the EZ81s will be working somewhere near their upper voltage rating, I would wire each valve as a half wave rectifier with anode sharing resistors as you suggest.

PY82s work well in pairs as full wave rectifiers if you can power the heaters. They're a lot cheaper...

Leon.

[G6Tanuki]

I suppose anode-to-anode flashover in my suggested circuit could be a problem; though the EZ81 is rated for 350V HT and we're only proposing a 325-0-325V supply.

We've already got the EZ81s [and the 6L6s and the couple of 6SN7/VT231 needed for the audio amp/phase-splitter]; if we were doing this and having to *buy* rectifiers I would be really-suggesting that we use silicon rectifiers instead!

[kalee20]

I'd be inclined to use each EZ81 as a full-wave rectifier. But have a 150 ohm resistor in series with each anode.

If one of the valves fails, you still have FW rectification by the other one. In your method, you'd have HW.

Resistors in the anodes should be equivalent to a single resistor in the cathode, however if the valve goes soft and you have anode-anode flashover, a single resistor in the cathode won't provide any protection. Whereas anode resistors will limit the current - and if appropriately chosen, might even go O/C for you saving the transformer!

[Robert Gribnau]

Originally Posted by Leon Crampin

...
PY82s work well in pairs as full wave rectifiers if you can power the heaters. They're a lot cheaper...
...

The maximum input voltage of the PY82 is only 250 Veff (so 2 x 250 Veff for a pair connected as full wave rectifier) which is too low for most amplifiers with 6L6's in push-pull.

A pair of EY84's seems a better candidate to me for such a full wave rectifier:

https://frank.pocnet.net/sheets/030/e/EY84.pdf

[PJC58-Hythe]

This subject has been discussed before https://www.vintage-radio.net/forum/...d.php?p=816603 and this might give some additional insight. Personally I like the example of the Klemt M40 shown in the thread, although I think I would be putting resistors in each of the anodes.

[G6Tanuki]

Given that the intended user of this amplifier is a vintage-synth afficionado and not really interested in the minutiae of current-sharing or Thevenin equivalents, and not wanting to 'frighten the horses' with anything involving *equations*, I've decided that we will keep it simple and strap the two anodes of each EZ81 together via 120 Ohm current-sharing/limiting resistors, and then treat them as a pair of 'composite diodes' either side of the transformer centre-tap.

Nice big metalclad resistors bolted to the chassis should run cool.

Then into 100uF of smoothing; the EZ81 specified a max of 50uF but that is presumably per-anode; with the two anodes paralleled doubling the smoothing should be safe.

[The transformer has a 6.3V@2.5A, with a tap at 5V, designated for rectifier use, which we will leave floating, so there's no problems with heater-cathode voltage limitation on the EZ81s.]

[murphyv310]

My stereo 3-3 amp has a single 350v HT winding from which I have a capacitor dropper to a bridge rectifier and then onward to a strapped anode EZ81, the value of the dropper cap was worked out by substituting values to give 310v HT once the amp has settled after 15 minutes.
After about five years use this was 305v last year. The EZ81 is only there for slow HT build up, smoothing is after the EZ81. I also incorporated a fuse in the circuit. Works perfectly for me.

[G6Tanuki]

Noted. Given that we are working with pre-existing valves and mains-transformer, I don't want to complicate things with using only one half of the mains transformer and a bridge; sure - personally if I was doing this from first principles I'd be using a single-secondary HT transformer into a silicon bridge and forgoing any lossy valvery in the HT-path-to-the-smoothing-caps.

But we work with what we have. My aim is for something like 35-40W from the push-pull 6L6 pair; we're going for grounded-cathodes and a separate negative-bias supply for the grids [with pots to adjust the grid bias individually for eadh valve].

Ssssh!! Don't mention that we will be using semiconductors [a 7924 three-terminal -ve voltage regulator with a Zener in the earth-path to get extra -ve voltage].

[murphyv310]

Absolutely
The mains transformer I have is quite a beast with a single 350v winding at 200ma and 6.3v @ 8a. With just a single heater supply I thought it wise to incorporate a 150ma fuse prior the the EZ81 Anodes which are strapped and no limiting resistors. Of course the valve is fed from a DC source and limiting via the capacitor prior to the silicon bridge.

[kalee20]

Originally Posted by G6Tanuki

Given that the intended user of this amplifier is a vintage-synth afficionado and not really interested in the minutiae of current-sharing or Thevenin equivalents... I've decided that we will keep it simple and strap the two anodes of each EZ81 together via 120 Ohm current-sharing/limiting resistors, and then treat them as a pair of 'composite diodes' either side of the transformer centre-tap.

That'll work!

(Actually, the owner won't need to be interested in current-sharing and Thevenin equivalents... it's all transparent to him/her).

But what's the reason for each EZ81 as a full half-wave, as opposed to each being half the full-wave, as my post #4, which doesn't involve any extra wiring or resistors?

Originally Posted by G6Tanuki

Then into 100uF of smoothing; the EZ81 specified a max of 50uF but that is presumably per-anode; with the two anodes paralleled doubling the smoothing should be safe.

Agree! (To be pedantic, it's the cathode rather than the anode which determines the rating... but as each anode has its own area of cathode, it's valid.

Originally Posted by G6Tanuki

The transformer has a 6.3V@2.5A, with a tap at 5V, designated for rectifier use, which we will leave floating, so there's no problems with heater-cathode voltage limitation on the EZ81s.

That's the only thing that niggles me, though it's a personal pet-hate, floating heaters! If the transformer winding is thus intended, its insulation will be impeccable, so you might as well common the heater winding to cathode and avoid the possibility of it drifting to an undefined voltage with possibly a micro-breakdown flashover to something from time to time.

[ms660]

Also don't forget that the total circuit resistance must be high enough to limit the diodes peak transient current to 1.8 Amps max. as per the valve data:

https://frank.pocnet.net/sheets/010/e/EZ81.pdf

The total circuit resistance is the resistance of the surge limiter plus the resistance of one half of the secondary winding (which should include the resistance reflected by the primary) plus the resistance of the diode at the maximum allowable peak transit current, as an approximation the latter will be somewhere around 60 Ohms.

Lawrence.

[G6Tanuki]

Here's what I have finally come up with.

The pair of EZ81 have their anodes fed from the ends of the HT transformer secondaries, via 120 Ohm load-sharing resistors shown as R1. They will be 5-Watt-rated metalclad types. [the HT transformer secondary windings have a DC resistance of between 77 and 91 Ohms, so provide additional peak-current-limitation. Purists would no doubt recommend increasing the load-sharing resistors' values in the lower-resistance half of the transformer secondary; I'm not so pedantic].

C1, the first capacitor, is 100uF; R2 is 100 Ohms, again metalclad; the feed from it goes to the centre-tap of the 6L6 output-transformer. C2 is 180uF, R2 is 1KOhm; they feed the screen-grids of the two 6L6. R3 is 5K6, C3 is 470uF to feed the low-level audio stages with a reduced voltage [260V] and additional decoupling.

LP1 is a 12-volt 3-Watt MES bulb which acts both as a HT fuse and a pilot-light.

[Lucien Nunes]

LP1 is a 12-volt 3-Watt MES bulb which acts both as a HT fuse and a pilot-light.

And a trap for the unwary who are not expecting one end of a 12V pilot light in a transformer-fed circuit to be hundreds of volts from deck if it blows.

[ms660]

Originally Posted by ms660

The total circuit resistance is the resistance of the surge limiter plus the resistance of one half of the secondary winding (which should include the resistance reflected by the primary) plus the resistance of the diode at the maximum allowable peak transit current, as an approximation the latter will be somewhere around 60 Ohms.

Whoops!...47 Ohms (approx)

http://www.r-type.org/pdfs/ez81-1.pdf

Lawrence.